Map the OpenClaw-RL paper and repo into a bounded Ratchet next-state / directive-correction audit slice without claiming the OpenClaw-RL training stack is ported.
Use this skill when we want to process OpenClaw-RL as source material and make one honest Ratchet-native slice from it without pretending we imported its RL runtime.
next-state, directive correction, and async improvement ideas onto existing Ratchet seamswitness-recorderruntime-state-kernelbounded-improve-operatora2-skill-improver-readiness-operatora2-skill-improver-first-target-proof-operatorWhen no explicit path is supplied, write:
system_v4/a2_state/audit_logs/A2_NEXT_STATE_SIGNAL_ADAPTATION_AUDIT_REPORT__CURRENT__v1.jsonsystem_v4/a2_state/audit_logs/A2_NEXT_STATE_SIGNAL_ADAPTATION_AUDIT_REPORT__CURRENT__v1.mdsystem_v4/a2_state/audit_logs/A2_NEXT_STATE_SIGNAL_ADAPTATION_AUDIT_PACKET__CURRENT__v1.json